Page 59 - Read Online

Lyons et al. Cancer Drug Resist 2021;4:745-54 https://dx.doi.org/10.20517/cdr.2021.37 Page 749

broad diversity of human PDO models available will massively enrich research efforts and possibilities to
validate ADC effects in various disease models and thus enable the stratification of the patients who will
most likely benefit from treatment with a specific ADC molecule.

ADVANCES IN GENETIC MANIPULATION
A critical aspect of ADC development is to stringently test and understand the specificity of binding to the
target antigen. This knowledge is ultimately key for successful clinical translation and identifying which
individual patients are most likely to benefit from treatment with the ADC.

The specificity of the antigen is often demonstrated experimentally using different, non-genetically matched
tumor cell lines that differentially express the antigen or a pre-treatment block with non-labeled antibody.
These are acceptable methods, but recent advances in genetic manipulation, namely the widespread
adoption of CRISPR (clustered regularly interspaced short palindromic repeats), have greatly facilitated the
potential adoption of significantly higher experimental standards.

An ideal experimental scenario to test antigen specificity in vivo would be to employ paired preclinical
tumor models, identical in every way except for target antigen expression. Tumors plus or minus antigen,
developing on contralateral flanks or in matched experimental cohorts, would elegantly demonstrate that
observed ADC accumulation at a tumor is specific and not due to off-target interactions or simple passive
accumulation via leaky tumor vasculature and the EPR effect.

Traditionally, the only available approach to precisely knock-out the expression of a target gene relied upon
[16]
homologous recombination between a targeting vector and endogenous allele . The approach was
technical, inefficient, and time consuming to employ and was seldom performed in any other context than
in the targeting of murine ES cells to produce genetically modified mice. The subsequent discovery of RNAi
[17]
and the ability to easily and specifically knock-down gene expression with shRNA or miRNA was
[18]
transformative by rapidly facilitating specific and significantly reduced levels of gene expression in
practically any eukaryotic cell line of interest. This approach is still valid today, however gene knock-down
by RNAi is frequently not 100% and so the target antigen is still expressed to some extent.

The relatively recent discovery, rapid development, and widespread adoption of CRISPR technology has
completely revolutionized our ability to precisely modify the genome . Among the many documented
[19]
applications, it is now relatively straightforward and efficient to generate such matched tumor model pairs,
plus or minus the expression of antigen. Briefly, CRISPR introduces precisely targeted double-strand breaks
in the genome, which are typically imperfectly repaired by non-homologous end joining. This repair
process often results in the microdeletion of one or more nucleotides at the DSB. Accordingly, bi-allelic
frame-shifting mutations can be readily introduced into the specific coding sequence of a gene of interest,
effectively knocking out its expression.

On occasion, knock-out of gene expression is poorly tolerated by the targeted cell, significantly affecting cell
fitness or causing phenotypic drift, such that the genetically paired cell lines are no longer a good match
biologically. In those circumstances, inducible transgene technology can be employed (e.g., doxycycline
[20]
inducible expression ) to limit the amount of time between antigen knock-out and antibody affinity assay.
Further, inducible gene expression can introduce or restore the expression of an antigen in a cell line that
the ADC does not otherwise recognize. For example, the expression of a tumor-specific antigen could be
readily introduced into a “normal” and non-expressing organoid cell line (as mentioned in Section i);
implanted cells in non-induced mice will not express the antigen and so lack affinity for the ADC, whereas